This year we invited extraordinary Professor Peet Strydom to write a series of guest posts about economic theory. In this second post he considers monetary policy after the financial crisis, specifically quantitative easing (QE):
MONETARY POLICY AFTER THE CRISIS: QUANTITATIVE EASING
Prior to the financial crisis that started in August 2007 monetary policy was, generally speaking, conducted within a one instrument one target framework. The financial crisis affected the liquidity position of banks adversely and this liquidity constraint transferred the depressing effect of the financial crisis to the real economy and the great recession started in 2008. Central banks were quick to supply the financial sector with financial liquidity while restraining the recession through lower interest rates or an expansionary monetary policy. Unfortunately monetary policy reached its full impact quite rapidly as short-term interest rates reached the zero lower bound. Conventional monetary policy came to a halt and central banks in the US, the UK and the ECB embarked on balance sheet expansion as a new policy measure.
We describe unconventional monetary policy in terms of quantitative easing (QE) in the following section. Thereafter we review the assessment of the success of QE as seen by New Keynesian economics. We extend our exposition by discussing assessments that followed alternative methodologies.
1. Quantitative easing as an unconventional monetary policy instrument
As indicated by Joyce et al (2012) unconventional monetary policy takes several forms but the more common one involves large increases in central banks’ balance sheets. The Federal Reserve also introduced Operation Twist with a view of influencing non-standard interest rates. These transactions were merely a switching of assets of different maturity in the financial markets without affecting the size of the Fed’s balance sheet.
QE is the most prominent form of unconventional monetary policy. In the UK the central bank conducted QE primarily by buying government bonds from the non-bank sector. The Fed concentrated on US Treasury bills but also bought debt instruments from government agencies such as mortgage backed securities. The ECB expanded its balance sheet largely through repo operations. It supported EU banks through financial sector collateral as opposed to government bonds. The ECB addressed the liquidity problem of EU banks while the Fed and the BoE tried to affect the yields on a wide range of assets, primarily those associated with the financing of households and companies. Although central bank financial market activities affected the yields on government bonds they concentrated their activities on private sector security yields to support private sector funding, also referred to as credit easing. Outright Monetary Transactions (OTM) is a particular ECB instrument to exert downward pressure on long-term yields of Eurozone government bonds. Such transactions are subject to strict “conditionality” by the IMF, the ECB and the European Commission. No country has, as yet, applied for ECB support under OTM rules.
The main purpose of QE is to encourage portfolio switching in the private sector as suggested by Tobin (1958, 1961) i.e. in terms of a portfolio balance approach. The effectiveness of balance sheet policies is dependent on the success of transferring portfolio switching effects to the real sector.
The switching or substitution of assets can be explained by considering a purchase of long-term government bonds from non-bank private owners. Usually these are insurance companies or pension funds because they are the principal holders of such assets. The direct effect of such purchases is that the supply of such bonds to the private sector declines. In order to maintain the desired demand for long-term bonds in the private sector some of these proceeds of the sale are earmarked for the purchasing of more risky higher yield private sector securities. This implies an increased demand for corporate bonds and equities that exert upward pressure on their prices. Rising asset prices attract corporates with healthy balance sheets and cash balances to participate in asset transactions, securing trading volumes. The rising prices give households capital gains which boost their wealth. The wealth effect is likely to stimulate demand and GDP through the well-known wealth effects of the macroeconomic consumption function. Rising prices of private sector assets enable companies to secure funding through easier credit conditions. Furthermore, rising equity prices open the possibility of corporate funding via the capital market. Moreover, merger and acquisition possibilities become possible. These developments are likely to be favourable to investment spending as they are aimed at constraining and reversing recessionary effects in the real sector.
The switching activity in bond markets can be refined in terms of particular processes or behaviour patterns that constitute transmission channels. The most extensive elaboration on these channels is by Krishnamurthy and Vissing-Jorgensen (2011). The extent of asset switching is determined by the behaviour of market participants and also by the category of assets that have been targeted by the central bank.
The substitution or switching activities mentioned above are not only evident between markets but also within markets. This means that financial markets are segmented. One explanation of segmentation is the preferred-habitat view by Vayanos and Vila (2009). They emphasised the preference of certain market participants for specific maturities as originally suggested by Culbertson (1957). Examples are specific maturity preferences by pension funds as opposed to the preferences by asset managers for own account and banks. In this exposition arbitrageurs are important in transmitting changing supply and demand conditions from a particular habitat to the general term structure of interest rates. Arbitrageurs prevent extreme segmentation in bond markets while they change yield patterns through carry trade. When short-term yields rise, they sell bonds to invest in higher short-term yields. Bond prices fall and long yields rise. As opposed to this carry trade “roll-up”, there is the carry trade “roll-down” when short-term yields are low. Arbitrageurs borrow at the short end and invest in bonds and bond prices rise and long yields fall. As indicated by Vayanos and Vila (2009), monetary policy interventions affect interest rates that generate wealth transfers to arbitrageurs through changing risk premia.
Gagnon et al (2010) paid explicit attention to the importance of the risk premium in capital markets. The reduced supply of riskier long-term assets, owing to central bank intervention, implies a fall in the risk premium of holding such assets and their yields decline. Gagnon et al (2010) associate falling interest rates with lower risk premia as opposed to expectations about future short-term interest rates.
2. New Keynesian assessment of QE
New Keynesian macroeconomics adheres to Keynesian rigidities that counter act the typical general equilibrium adjustment process as subscribed to by new classical macroeconomics. This branch of Macroeconomics acknowledges the existence of disequilibria such as involuntary unemployment while accepting a role for stabilisation policy. This intellectual framework with its special tool, Dynamic Stochastic General Equilibrium (DSGE) models, as for instance introduced by Smets and Wouters (2007), has already made substantial progress in the development of different types of DSGE models. These models, as a rule, predict that monetary shocks are likely to have small and non-persistent effects on the real sector.
An interesting example of a policy analysis in terms of this framework has been presented by Eggertsson and Woodford (2003). The authors investigated the effectiveness of monetary policy in the event of a liquidity trap, or what is also referred to as the zero lower bound interest rate. The object of the analysis is the Japanese experience that resulted in the long-term Japanese stagnation period. The authors concluded that quantitative easing had little chance of success in Japan. Although this conclusion is correct because of several reasons that will not be discussed here, one has to keep in mind that the Japanese case with its zombie banks at the time is not fully comparable with that of present day US conditions. The authors provide a very negative assessment of the effectiveness of QE as a policy instrument.
Looking at the analytical framework of the Eggertsson and Woodford DSGE model one cannot expect it to give a positive assessment of QE. It relies on the micro foundations of a utility maximising representative agent facing homogeneous pricing in terms of the Calvo proposition. Financial assets are considered to be perfect substitutes and only a single interest rate is applied. These restrictive assumptions appear to be in support of homogeneous bond markets. In the real world these markets are non-homogeneous owing to different maturities and dissimilar liquidity requirements as discussed by Culbertson (1957). As has been indicated above, QE relies on the possibility of portfolio adjustments which means that the intellectual framework of this model is not very appropriate to assess monetary policy effects under modern institutional arrangements. Market participants do not consider financial assets of a different risk and maturity profile as identical. Moreover, the micro foundations that rely on representative agent behaviour, coupled with the ad hoc homogeneous pricing behaviour are not very helpful in analysing real world behaviour. The ad hoc homogeneous pricing behaviour in DSGE models had a critical reception in the literature as argued by Alvarez (2008) as well as Creamer, Farrell and Rankin (2012).
In a more recent paper Cúrdia and Woodford (2010) applied a closed economy DSGE model with several interest rates and intermediation without prior knowledge on the financial position of borrowers. Despite these important additions to the DSGE model the authors’ assessment is that QE does not appear to be a helpful monetary policy tool, even under zero lower bound conditions. In similar vein Chen, Cúrdia and Ferrero (2012) modified a typical Smets and Wouters DSGE model to investigate the macroeconomic effects of unconventional monetary policy in the US. A novelty of their approach is the segmentation of the capital market; distinguishing different agents that do not all can take advantage of arbitrage opportunities. The analysis, therefore, escapes the Wallace (1981) irrelevance result. The latter denies macroeconomic effects coming from changes in government asset portfolios under restrictive conditions of perfect foresight and perfect competition. The outcome of these DSGE expositions is that large scale asset purchases by central banks have very modest macroeconomic effects.
3. Non-mainstream assessments of QE
Several empirical studies have been conducted on the effectiveness of central bank balance sheet policies including the Federal Reserve, the BoE and the ECB. As indicated by Joyce et al (2012), the empirical evidence on the effectiveness of unconventional policy instruments is primarily based on central bank studies. These results are subject to certain qualifications but it lies beyond the aim of this document to elaborate on these debates. At this juncture it suffices to mention that the results are not independent of special circumstances in the different countries. Moreover, the results are likely affected by spill over effects between countries that will become evident in an open economy analysis. Moreover, supporting policies such as fiscal measures within a country will affect the outcome of the results.
The conclusions indicated above are echoed by Joyce et al (2012) in their critical review of the empirical research on QE in the US and the UK. In respect of the ECB the conclusion is that although the ECB concentrated primarily on maintaining Eurozone bank liquidity, the actions by the ECB were effective in supporting the Eurozone economy in terms of the performance of industrial production and employment. There have been reports in the international financial press regarding diminishing returns to QE but, as yet, there is no firm evidence on this.
At this juncture it is important to emphasise the empirical results that were obtained on the effectiveness of QE through an event-study methodology. This approach examines interest rate changes associated with official communications regarding asset purchases. The cumulative change is taken as a measure of the overall effect. The challenge is to include all the official communications regarding a policy event while assuming that they are the prime trigger in market adjustments. The time framework of the event, also referred to as the window, should be wide enough to capture the market effects and not too wide to be distorted by outside developments. The major advantage of an event-study is that it is conducted in calendar time as opposed to logical time of new Keynesian DSGE models.
The outcome of event-studies confirms the significance of QE effects, primarily regarding the term structure of interest rates. A large number of empirical channels were investigated in terms of US data by Krishnamurthy and Vissing-Jorgensen (2011). Their research emphasised the significance of signalling effects associated with preliminary announcements by the monetary authority. Gagnon et al (2010) emphasised the importance of the risk premium. Joyce and Tong (2012) considered several channels for UK data while confirming significant and persistent effects of QE transactions on yields.
The empirical evidence raises concerns about the size and duration of QE actions. There is no firm evidence to guide practical decision making on these aspects. It is probably fair to observe that it is more important to rely on evidence regarding the direction and effectiveness of policy measures while accepting error margins on size and duration. There is also widespread concern about the high level of bank reserves in the wake of QE and the likely distorting effects on the interbank money market. In similar vein, one could signal potential inflationary effects of these bank reserves.
Central banks have, nevertheless, assured market participants of an orderly reversal of large asset purchasing activities. Moreover, Siegel (2013) argued convincingly in favour of reserve requirements as an effective instrument to supplement other measures in preventing disruptive and inflationary effects. One could add weight to this argument on the strength that the vast liquidity creation by central banks is primarily circulating within the banking sector. All in all, we confirm empirical support for the effectiveness of QE transactions, primarily by research methodologies applying calendar as opposed to logical time.
The rapid escalation of recessionary effects following the financial crisis exerted pressure on central banks for a speedy expansionary monetary policy. The zero lower bound interest rate level of a liquidity trap signalled the limits of conventional monetary policy. This condition was becoming worse in view of the fact that, in the major developed countries, fiscal policy had very little operational strength owing to large and expanding government debt, coupled with budget deficits. Given these constraints, Central banks introduced unconventional monetary policies. New Keynesian macroeconomics appears to be critical of the effectiveness of these measures and rule out sustainable favourable effects on the economy. In contrast to this conclusion it appears that non-mainstream macroeconomic research do signal favourable effects coming from QE at the zero lower bound. We have argued that the micro foundations of New Keynesian economics are inappropriate to analyse the effectiveness of QE at the zero lower bound. We cannot confidently accept the outcome of this research on the effectiveness of QE and for now we are in support of the research outcome of non-mainstream economics.
ALVAREZ, L.J. (2008) What do Micro Price Data Tell Us on the Validity of the New Keynesian Phillips Curve? The Open-Access, Open-Assessment E-Journal, 2 (19 June):1-36.
CHEN, H., CúRDIA, V., AND FERRERO, A. (2012) The Macroeconomic Effects of Large-scale Asset Purchase Programmes, Economic Journal, 122(564):F289-F315.
CREAMER. K., FERRELL, G. AND RANKIN, N. (2012) What Price-level Data can tell us about Pricing Conduct in South Africa, South African Journal of Economics, 80(4):490-509.
CULBERTSON, J. (1957) The term Structure of Interest Rates, Quarterly Journal of Economics, 71(4):485-517.
CúRDIA, V. AND WOODFORD, M. (2010) The Central-Bank Balance Sheet as an Instrument of Monetary Policy, NBER, Working Paper No. 16208.
EGGERTSSON, G. AND WOODFORD, M (2003) The Zero Bound on Interest Rates and Optimal Monetary Policy, Brookings Papers on Economic Activity, 1:139-233.
GAGNON, J., RASKIN, M., REMACHE, J. AND SACK, B. (2012) Large-scale Asset Purchases by the Federal Reserve: Did they Work? Federal Reserve Bank of New York, Staff Reports, No. 441, March.
JOYCE, M.A.S. AND TONG, M. (2012) QE and the Gilt Market: A Disaggregated Analysis, Economic Journal, 122(564):F348-F384.
JOYCE, M., MILES, D., SCOTT, A. AND VAYANOS, D. (2012) Quantitative Easing and Unconventional Monetary Policy – An Introduction, Economic Journal, 122(564):F271-F288.
KRISHNAMURTHY, A. AND VISSING-JORGENSEN, A. (2011) The effect of Quantitative Easing on Interest Rates: Channels and Implications for Policy, Brookings Papers on Economic Activity, 2:215-265.
SIEGEL, J. (2013) Marketing Insight: Fed Must Revert to Tested Reserve Policy, London: Financial Times, 27 March.
SMETS, F. AND WOUTERS, R. (2007) Shocks and Frictions in US Business Cycles: A Bayesian DSGE Approach, American Economic Review, 97(3):586-606.
TOBIN, J. (1958) Liquidity Preference as Behavior towards Risk, Review of Economic Studies,25:65-86.
TOBIN, J. (1961) Money, Capital and other Stores of Value, American Economic Review, 51(2):26-37.
VAYANOS, D. AND VILA, J-L (2009) A Preferred Habitat Model of the Term Structure of Interest Rates, NBER Working Paper No. 15487.
WALLACE, N. (1981) A Modigliani-Miller Theorem for Open-Market Operations, American Economic Review, 71(3):267-274.